Identification of 17 Pseudomonas aeruginosa sRNAs and prediction of sRNA-encoding genes in 10 diverse pathogens using the bioinformatic tool sRNAPredict2

Abstract

sRNAs are small, non-coding RNA species that control numerous cellular processes. Although it is widely accepted that sRNAs are encoded by most if not all bacteria, genome-wide annotations for sRNA-encoding genes have been conducted in only a few of the nearly 300 bacterial species sequenced to date. To facilitate the efficient annotation of bacterial genomes for sRNA-encoding genes, we developed a program, sRNAPredict2, that identifies putative sRNAs by searching for co-localization of genetic features commonly associated with sRNA-encoding genes. Using sRNAPredict2, we conducted genome-wide annotations for putative sRNA-encoding genes in the intergenic regions of 11 diverse pathogens. In total, 2759 previously unannotated candidate sRNA loci were predicted. There was considerable range in the number of sRNAs predicted in the different pathogens analyzed, raising the possibility that there are species-specific differences in the reliance on sRNA-mediated regulation. Of 34 previously unannotated sRNAs predicted in the opportunistic pathogen Pseudomonas aeruginosa, 31 were experimentally tested and 17 were found to encode sRNA transcripts. Our findings suggest that numerous genes have been missed in the current annotations of bacterial genomes and that, by using improved bioinformatic approaches and tools, much remains to be discovered in 'intergenic' sequences.

Figure 1

Detection of novel sRNAs by northern analysis. Total RNA was extracted from cultures of P.aeruginosa strain PAO1 grown in LB to exponential phase (first lane in each blot) or stationary phase (second lane in each blot). Blots were hybridized to radiolabeled DNA oligonucleotide probes and then exposed for varying times; thus the relative intensities of the signals do not correspond to the relative abundance of each sRNA. The approximate positions of size standards are shown on the left. Boxes are included to highlight the major species observed in each blot.

Figure 2

The accuracy (closed diamond) and sensitivity (closed square) of the predictive search increases and decreases, respectively, as the BLAST stringency is increased. The accuracy corresponds to the percentage of sRNAs predicted at the indicated BLAST stringencies that were confirmed. The sensitivity corresponds to the percentage of all 23 experimentally confirmed P.aeruginosa sRNAs that were predicted at the indicated BLAST stringencies.

Figure 3

Venn diagram showing the number of novel sRNAs confirmed per the number of novel sRNAs predicted based on conservation between P.aeruginosa and the three BLAST partner species used for comparison.